Draven AI API

Draven AI exposes an OpenAI-compatible API for chat completions, plus dedicated endpoints for image generation, embeddings, text-to-speech, rerank, and web search. If you've ever called the OpenAI API, or used a client built for it, you already know how to use most of this one.

https://api.dravenai.lat/v1

Every response — success or error — is shaped exactly like the OpenAI API: the same chat.completion object, the same chat.completion.chunk streaming events, and the same { "error": { "message", "type", "param", "code" } } error envelope. This means the official OpenAI SDKs, the Vercel AI SDK, LangChain, LiteLLM, Open WebUI, LibreChat, and plain curl all work against Draven AI without any custom integration code.

How these docs are organized

• Docs — the concepts you need regardless of endpoint: authentication, rate limits, the free plan, and errors.
• Getting Started — a quickstart, how to generate an API key, and how to migrate an existing OpenAI integration.
• API Reference — one page per endpoint or behavior: models, chat completions, streaming, structured responses, reasoning models, vision, image generation, embeddings, text-to-speech, rerank, web search, and usage & logs.
• Agents & Integrations — wiring Draven AI into agent frameworks like OpenClaw and Hermes Agent, and Coding Tools like Cursor and Cline.

At a glance